Electricity of Drops. 357 



solutions of eosine or fluorescene. We have seen, however, 

 that at high temperatures the electrification of a solution 

 of rosaniline changes its sign ; and the state of the outer 

 layers when exposed to intense light may very well be more 

 analogous to that of the liquid at a high temperature than at 

 a low one. 



Again, the formation of large drops of water by the impact 

 of smaller ones is analogous to the splashing of the water- 

 drops against a wet surface, and is likely to give rise to analo- 

 gous electrical effects on a smaller scale. If this is so, the 

 large drops of rain which frequently accompany thunderstorms 

 may be to some extent rather the cause than the effect of the 

 storm. I believe that this has been suggested by Sir G. G. 

 Stokes. 



The difference in the behaviour of different gases with 

 reference to the two electricities is very conspicuous in these 

 experiments, oxygen and chlorine acquiring a charge of 

 negative electricity when under the same circumstances 

 hydrogen acquires a positive charge. This suggests that the 

 energy possessed by an atom of hydrogen, for example, when 

 charged with a unit of positive electricity is not the same as 

 that possessed by the same atom when charged with a unit of 

 negative electricity ; or, as v. Helmholtz expresses it, the atoms 

 of various substances attract the two electricities with different 

 intensities. If the atoms possess this property they will tend 

 to acquire definite atomic charges, and thus tend to have a 

 definite valency. We can see this if we remember that the 

 likelihood of the formation of a chemical compound is con- 

 ditioned by the changes in the energy which accompany the 

 formation of the compound. We shall for the sake of clearness 

 confine our attention to the changes which go on in the poten- 

 tial energy. Any chemical change will tend to go on if it is 

 accompanied by a decrease in the potential energy, and will tend 

 to be reversed if it is accompanied by an increase in that energy. 

 Let us consider the case of an element the potential energy 

 of whose atom is diminished by Q when one unit of positive 

 electricity is communicated to it. If C is the electric capacity 

 of the atom, then, when the atom has a charge of n units of 

 positive electricity, its electric potential energy is 



Thus the charging; of the atom will result in a diminution of 

 the potential energy until, supposing the unit of electricity 

 to be indivisible, the charge on the atom is the integer just 

 less than CQ : when the charge exceeds this an increase in 

 Phil Mag. S. 5. Vol. 37. No. 227. April 1894. 2 B 



